The long-term objective of this project is to identify nutritional factors that can lower C-reactive protein and lipid peroxidation. C-reactive protein (CRP), a marker of inflammation, has been shown to be "remarkably consistent" in its association with cardiovascular disease, as well as with diabetes and other conditions. Oxidative damage biomarkers have been associated with the etiology of numerous disease conditions, including atherosclerosis, diabetes, COPD and others. Both CRP and lipid oxidation have been shown to have a direct (not correlational) effect on vascular and other tissues, including effects on smooth muscle contractility, induction of intercellular and vascular cell adhesion molecules, enhancement of complement activation, and other mechanisms. We hypothesize that some of the harmful effects of obesity, and of post-menopausal status, are attributable to fat induced CRP and lipid oxidation. We propose to test whether antioxidant supplements can lower CRP and lipid oxidation. In a previous randomized trial we found that antioxidant supplementation significantly lowered CRP and lipid peroxidation. That study also found higher baseline CRP and lipid peroxidation in women compared to men, and in obese persons compared to those of normal weight. The previous intervention study did not measure body fat (except as BMI), assess menopause status, or intervene on persons who neither smoked nor were passively exposed. We now propose a study in nonsmokers, to test whether antioxidant supplements can lower CRP and oxidative damage. In the baseline data prior to intervention, we will determine whether the higher oxidation and CRP in women and overweight persons can be confirmed in persons with no smoke exposure, after adjustment for covariates. We will conduct a randomized placebo-controlled trial to test the following specific aims: 1. To determine whether antioxidant supplementation can lower C-reactive protein (CRP) and lipid peroxidation. 1a. To determine whether one antioxidant is significantly more effective than the other. 1b. To determine whether treatment effects differ by gender or body fat 2. To determine whether there are gender or body fat differences in baseline CRP or in baseline lipid peroxidation, in nonsmokers, independent of other covariates. 2a. In women, to determine whether postmenopausal status is associated with higher baseline CRP or lipid peroxidation, after control of covariates.